LiDAR点云辅助的高分辨率卫星影像建筑物地面阴影提取研究

With the increasement of spatial resolution in remote sensing imagery, the building ground shadow is becoming an issue for image processing and analysis. This paper proposes a method for automaticextraction of building ground shadow from high resolution satellite imagery with the help of the LiDAR cloud points. First, the building ground shadowregions in object space is determined by the imaging parameters of the high resolution satellite imagery and the LiDAR cloud points. Then, the ground shadow regions are projected to image space by an improved remote sensing satellite geometric positioning model, and the building ground shadow on the remote sensing imagery could be obtained. Finally, the experiment is carried out for high resolution QuickBird satellite imagery in Shanghai, China, together with the accuracy analysis, demonstrating the effectiveness of the proposed method.